Casper da Costa-Luis

Your python3 also needs to have numpy installed (if you `pip install "nipet>=2"` this would be done for you)

Yes but I presume you're using some sort of custom CMake wrapper so you need to `find_package()` first... have a look at https://github.com/NiftyPET/NIPET/blob/master/niftypet/CMakeLists.txt

Wait hang on... your output shows you're picking up numpy. My guess is it's just really old. Try `pip install --upgrade --user numpy`

Curious. - NumPy: >=1.14 should work. Just tested on python=3.6.12 numpy=1.19.5 - CUDA: you should be able to set any of `CUDA_ROOT`, `CUDAToolkit_ROOT`, or `CUDACXX` env vars instead of adding...

ah right. Well this is what I've done on a clean Ubuntu machine (18.04 and 20.04): ```bash sudo snap install --classic cmake wget https://raw.githubusercontent.com/pypa/get-pip/master/get-pip.py python3 get-pip.py --user python3 -m pip...

Sorry just saw this... merged. FYI from the readme: > Note that the more recently created [snapcore/action-build](https://github.com/snapcore/action-build) and [snapcore/action-publish](https://github.com/snapcore/action-publish) effectively copy the `use_lxd` strategy might be worth using the officially...

`dynamic_ncols=True`?

one bar on multiple lines isn't (yet) supported, but you could always piggy back off nested bars with custom `bar_format`s to split the information across two bars.

duplicate of #630

Rather than some hack ([SO#7106417](https://stackoverflow.com/questions/7106417/convert-decimal-mark-when-reading-numbers-as-input)), maybe just use `tr , .`? ```sh LC_NUMERIC=pl_PL.UTF-8 seq 3 .1 5 | tr , . | tqdm --total 5 --update-to ```